Difference between revisions of "Signal Research Paper"

From Geospatial Analytics and Applications
Jump to navigation Jump to search
Line 33: Line 33:
 
==<div style="background:#143c67; padding: 15px; font-weight: bold; line-height: 0.3em; letter-spacing:0.5em;font-size:20px"><font color=#fbfcfd face="Lato"><center>RESEARCH PAPER</center></font></div>==
 
==<div style="background:#143c67; padding: 15px; font-weight: bold; line-height: 0.3em; letter-spacing:0.5em;font-size:20px"><font color=#fbfcfd face="Lato"><center>RESEARCH PAPER</center></font></div>==
  
You can view our research paper here.
+
You can view our research paper here -
  
  
 
<!-- End Body --->
 
<!-- End Body --->

Revision as of 22:23, 14 April 2019


Logo2.jpg

TEAM

PROPOSAL

POSTER

APPLICATION

RESEARCH PAPER

RETURN TO PROJECT GROUPS

ABSTRACT

Application of geospatial analytics on traffic accidents is imperative for urban development design to improve road safety. Targeting road traffic accidents specific to certain casualty groups is also essential for effective traffic management. However, clusters of traffic accidents on networks are likely to evolve overtime. This characteristic shows that traffic accidents are not isolated in time and space. Therefore, traffic accident analyses have to incorporate both time and space elements so that temporary resources by the traffic police could be allocated efficiently, or future road planning efforts by transport authorities could be implemented in the right direction. However, currently available web-based traffic collisions applications mainly focus on visualising traffic accident point patterns. A web-based geospatial analytics tool, SIGNAL, is then developed to address the need for an interactive network constrained spatio-temporal dashboard on traffic accidents. This paper aims to explore the use of network constrained spatio-temporal statistical techniques on traffic accidents data in Leeds, United Kingdom. Four key methods are employed to conduct analyses in SIGNAL. Firstly, Network Constrained Kernel Density Estimation is used to derive insights on traffic collision intensity patterns. To enable identification of statistically significant clusters, Network Constrained K-Function is incorporated. Lastly, Network Constrained Cross K-Function and Network Constrained Cross Pair Correlation are adopted for investigating correlation between traffic collision points and variables of interests, such as pedestrian crossings, motorway junctions and schools. The results obtained from our demonstration highlight key insights that could help transport authorities, traffic police and even business users better understand spatio-temporal clustering patterns and correlations of traffic accidents.

RESEARCH PAPER

You can view our research paper here -